EarthLink Says No Future for Municipal Wi-Fi

Glenn Fleishman writes "EarthLink dropped its final bombshell on city-wide Wi-Fi, saying that it wouldn't put more money in and was talking to their current deployed cities about the future. The company had won bids in dozens of cities, and then backed out of the majority of them before building or finalizing contracts a few months ago. The remaining towns they were building out, like New Orleans, Anaheim, and Philadelphia, will ostensibly be turned off unless local officials come up with scratch or a plan of their own. EarthLink pioneered the model of free-for-fee networks, where there would be no cost or upfront commitment from cities, and EarthLink would charge for network access. Apparently, you can't make money that way."

Comments

If thats not foul play, i dont know what is

[unity100]

Check that out. a company goes into many wifi bids, wins most of them, and then suddenly decides 'city wide is not worth it'.

thats foul play at its best. proxies, they are.

Re:If thats not foul play, i dont know what is

[Jeff DeMaagd]

Normally, I might agree, but if they did back out before the contract was signed, so the cities should be able to fall back to the next bidder in line. If they did sign the contract, then there should be penalties for not abiding by the contract.

Another problem is that WiFi just isn't very well suited for city-wide networks and it looks like these companies are finally figuring that out.

The consumer access points being cheap and virtually everyone's computer having a client-side adapter doesn't help the cost issue enough to help make it affordable to the users, unless the network rollout is charity work. You need to rent utility pole space on every pole or every other pole for APs, assuming there are utility poles, some cities have been pushing towards underground wiring. You'd also need to worry about getting power to the APs. For every block, one T1 wired network drop for connection to the internet. I don't think those access points are consumer units either. Even if they were, the weatherproof enclosures are expensive too. Then there would need to be maintenance. I just don't see a viable, affordable competitor coming out of that. To me, all that makes WiMax seem viable, relatively speaking. Maybe if someone like Canopy can make pocket EC/34 or USB network adapters, then I think Canopy would be a better alternative.

Re:If thats not foul play, i dont know what is

[div_2n]

First, you can legally use all eleven channels of 802.11b in the US. It just so happens that if you want to have perfect channel separation, channels 1, 6 and 11 are the way to go. I've sat in on a discussion about using 1, 4, 8 and 11 or some such combination and achieving acceptable separation provide the APs aren't on top of each other.

Regardless, using the three channels and 120 degree directional antennas to cover the full 360 is the most effective way. That isn't cheap. Even if you roll your own using a WRAP board or some such thing, last I checked you can't get a weather proof AP with all three channels and antennas for less than $1,000. That doesn't count labor.

Clearly I know absolutely nothing about this, but

[Gigiya]

When they make and win the bid, why aren't they legally obligated to follow through with it?

city-wide wifi has its uses

[erlehmann]

for example in the eastern part of germany, after reunification, there were lines in cities that could not be used for DSL. the german "freifunk" (literally "free wireless", both as in beer and as in speech) project managed to build some sizeable city mesh nets using a routing protocol known as OLSR [1,2].

just look in awe at the leipzig cloud [3]. also, try to imagine wireless cell phone / pda mesh nets (probably doable right now with openmoko).

[1] http://olsr.org/
[2] http://www.ietf.org/rfc/rfc3626.txt
[3] http://db.leipzig.freifunk.net/uptime/png/ -- careful, images is 3165x4206

Re:city-wide wifi has its uses

[fireboy1919]

using a routing protocol known as

You're kind of missing the point in this example. Citywide 802.11x is not going to work. The protocol can't handle it. It doesn't scale up to even 100 users at once.

The obvious solution is to use some other protocol.

Re:Unnecessary

[mustafap]

>I only walked a mile or so through mostly residential streets

Try using a car. I recently moved to a small sussex town, and found an open network in a few minutes when I needed internet access to find an estate agent. There are two open networks in my new street too.

City WiFi is for outdoors

[icepick72]

I've subscribed to OneZine city WiFi in Toronto, Canada and the signal degrades quickly as you move back from the street. The best signal is sitting on the sidewalk with your laptop ... with the homeless people. You also pick up a strong signal while driving of cycling on the street but ... not a lot of time to use it. Suffice to say I dropped it in favour of Starbucks Wifi/Bell Hotspots which have a stronger signal indoors. There are enough Starbucks around that I'm never without a connection.

Give us some spectrum and we'll make it happen

[troll -1]

I wonder if the problem with Municipal Wi-Fi, or Wi-Fi in general, is that companies like EarthLink are trying to operate with their hands tied behind their back with restrictive power limitations and limited frequencies while the FCC gives away large chunks of the best part of the spectrum to cell phone providers for millions of dollars who then nickel and dime us for every trivial service they can think of.

Perhaps he reason we don't have a ubiquitous and cheap wireless Internet and why TCP/IP mesh networks are *not* on the horizon for the 700MHz part of the spectrum is because the government insists on auctioning off a zero cost medium for mega bucks to legal monopolies who have no choice but to turn around and stick it their customers.

Maybe we need to stop thinking in terms of phone systems when we think about the spectrum and start thinking more in terms of extending the Internet. Just a thought.

Fucked again by Public Private Partnerships.

[Anonymous Coward]

This is yet another example of the general public being fucked over by a so-called P3: Public Private Partnership.

The idea is that instead of the municipal government setting up an organization to perform a specific project, they basically contract out the job to private firms. Supposedly this will lead to more economical and better quality service. Instead, what we've seen time and time and time again, is nothing but higher prices, and far shittier service.

Then we get cases like this, where the private interest just pulls out of the deal when it's no longer profitable for them. Of course, it doesn't matter that they've fucked over the community. A lot of the time these companies have little to no ties with the community they are servicing, so leaving the public there high and dry causes these private firms little grief.

volunteerism

[wikinerd]

Where is your volunteerism?? Why should you expect the government, a company, or anyone else to provide you with wifi service when you can roll out your own??

You are not consumers. You can be producers if you want. Just knock your neighbour's door and ask whether they would like to start a new wifi community network project with you. Connect your home wifis together, and if you find a lot of people to join in then you will have created your own network. Then buy a business plan fixed broadband service or a dedicated line (paid either by the community as a whole or by one richer member who can pay for it) and connect it to your wireless and your network will be connected to the Internet as well.

That simple. Yes, I know, the technology (WiFi) is not perfect and you can't transmit with too much power, but if everyone has a roof and the signal is sufficient from roof to roof, then you don't need anything else. The major difficulty is actually a social one (your neighbors may not understand what volunteerism is), but you should try to educate your neighbors and persuade them why they should join in.

Look what people from my city are doing: AWMN and also look at the photos and some other networks in existence worldwide.

The cage is open guys. You have unlicensed bands that you can use without a permit from FCC or other agency. You even can have RONJA if you like the optical way. You also have telephone lines, modems, and BBS software. Why you don't use all this technology to create free networks? Are you really trained to act only as consumers, expecting that for everything you need you should buy it from someone else? If you aren't happy with what is available, build your own!

Re:Clearly I know absolutely nothing about this, b

[eggboard]

It's a great point. In some cases, EarthLink had won a bid, but not started contract negotiations. In some cases, a contract was on the table, but not signed. In some cases, a contract was completed, but the city hadn't executed it (often, a mayor works out the details and a council approves it, and then a utility has to be involved to agree to pole uses).

Where a contract is in place, EarthLink will have to unwind its obligations. In Houston, it paid $5m for not starting the network. In Philadelphia, they will likely pay out millions to walk away.

Re:Unnecessary

[vertinox]

there are no so many home networks left open that one can just use whatever is nearby.

This is why my routers DNS entries redirect to a particular website with a particular image. You know which one.

802.11X is NOT Suitable for Last Mile

[nuintari]

I've said this a dozen times in related articles, but I'll say it again.

802.11 is the flat out, 100%, god awful, worst solution for last mile delivery. I work for a wisp that uses Canopy products, and we just laugh at the 802.11 competition. 802.11 performance degrades the more people you stuff on an access point. The limited channels, and the fact that they scream over each other forces competing networks to get into AMP powered frequency wars. The fact that only channels 1, 6 and 11 are clear from each other makes splitting an access tower to more than three 120 degree sectors pretty much impossible. And neighboring towers will interfere with each other. Oh, and because of how 802.11 does time sharing, essentially Ethernet collision detection with a few hacks on top, one nasty user can monopolize 95% of the available bandwidth for himself without much effort. And this is just my experience in the countryside, where we have few competitors to the last remnants of 802.11 we still have deployed. The reason no one can make money deploying 802.11 on a massive scale is because operationally speaking, it costs a bloody fortune to maintain.

Just because Moto's canopy is proprietary doesn't make it bad. They have been very good to us, old client radios work with newer access points, whenever a new generation of access points comes out, they have an awesome trade up deal that lasts for months, giving us plenty of time to give our customers the best speed available, without breaking the bank in one mass upgrade. There is an active 3rd party mailing list, that Moto monitors and responds to, an entire community of support from end ISPs, and mountains of documentation.

Do wireless right, make money, do it 802.11, and spend hours on the phone with irritated users who want to switch back to dialup.

Re:Unnecessary

[UbuntuDupe]

Time for analogy wars!

Do you just walk into random houses and use their computer or phone whenever you feel like it as well?

Yes, I do, if the computer/home has all of the the following attributes:

-There's a big sign outside that says "computer in here" (access points advertise their presence)
-There's an instruction set outside the house that says "To access the computer, rotate the knob on this door and push forward. Walk into home, then enter second room on right. Press power button, wait for authorization, and then use." (access points tell you how to use them)
-After pressing the power button, a message says "request for computer use received ... access granted" (access points must receive a request for use, and then grant permission)